J. Org. Chem. 1990,55, 1401-1406
1401
Scheme I
1) n-BuLi, MF. -78'
1) n-BuLi. THF. -78'
toluene, H,O
7
8
by suction. Crystallization from EtOH/EhO afforded colorless needles: 1.43 g (84%); mp 213-215 OC dec; IR (KBr) 1690,1680 cm-'; 'H NMR (DMSO-d6) 6 1.68-1.84 (m, 2 H), 2.05-2.18 (m, 2 H), 3.11 (dt, J = 12.6, 2.6 Hz, 2 H), 3.30-3.45 (m, 2 H), 4.13 (tt, J = 11.4, 3.5 Hz, 1H), 5.86 (s,2 H), 7.09-7.23 (m, 4 H), 7.37-7.51 (m, 2 H), 7.77 (d, J = 7.8 Hz, 1 H), 7.89 (d, J = 8.1 Hz, 1H), 8.62 C, 58.54; (br s , 1 H). Anal. Calcd for CZoHzoFN30CF3CO2H: H, 4.69; N, 9.31. Found: C, 58.55; H, 4.77; N, 9.29. [ 1-[(4-Fluorophenyl)methyl]-lH-benzimidazol-2-y1][ 1[2-(4-methoxyphenyl)ethyl]-4-piperidinyl]methanone (2). To a stirred, -78 "C, solution of 4 (1.13 g, 5.00 mmol) and dry THF (12 mL) under argon was added via syringe a 2.5 M solution of n-butyllithium in hexane (2.1 mL, 5.2 mmol). After 15 min a solution of 5 (1.39 g, 5.01 mmol) and dry THF (6 mL) was added dropwise via syringe. After 5-10 min, the cooling bath was removed, and after 30 min the reaction was quenched by the addition of saturated aqueous NH,Cl. The aqueous mixture was extracted twice with EtOAc. The EtOAc extracts were combined, washed with saturated aqueous NaCl, and dried over anhydrous Na2S04. The drying agent was removed by filtration, and the filtrate was evaporated at reduced pressure leaving an oil. Purification of this oil by flash chromatography (EtOAc) and crystallization from cyclohexane afforded off-white, matted needles: 0.90 g (38%); mp 109-111 "C; IR (KBr) 1680 cm-I; 'H NMR (CDClJ 6 1.71-1.88 (m, 2 H), 1.98 (br d, J = 11.9 Hz, 2 H), 2.24 (dt, J = 11.6, 2.5 Hz, 2 H), 2.54-2.62 (m, 2 H), 2.72-2.80 (m, 2 HI, 3-05 (dt, J = ii.6,3.2 HZ, 2 H), 3.78 ( ~ , H), 3 3.88 (tt, J = 11.5, 3.8 Hz, 1 H), 5.79 (s, 2 H), 6.79-6.86 (m, 2 H), 6.90-7.00 (m, 2 H), 7.06-7.17 (m, 4 H), 7.33-7.46 (m, 3 H), 7.87-7.96 (m, 1 H). Anal. Calcd for CBHgN302:C, 73.86; H, 6.41; N, 8.91. Found: c, 73.98; H, 6.48; N, 8-90. Alternate Preparation of 2. To a stirred, room temperature mixture of 8 (1.82 g, 4.03 mmol), KzCO3 (1.39 g, 10.0 mmol), and DMF (15 mL) was added 1-(2-bromoethy1)-4-methoxybenzene (0.87 g, 4.0 mmol). The reaction was then heated at 90 OC for 22 h. The reaction was poured into HzO,and the aqueous mixture was extracted with a 2:l mixture of EtOAc/toluene. The organic layer was washed twice with H20and once with saturated aqueous NaCl before being dried over anhydrous Na2S04. The drying agent was removed by filtration, and the filtrate was evaporated at reduced pressure leaving an oil. Purification of this oil by flash chromatography (EtOAc) and crystallization from cyclohexane afforded off-white, matted needles: 1.08 g (57%). This material was spectroscopically identical with the previous sample of 2. Anal. Calcd for CWH&N3O2: C, 73.86; H, 6.41; N, 8.91. Found: c , 73.73; H, 6.50; N, 8.90. 1 4 4 4l,l-Dimethylethyl)phenyl]-4-[4-[[1-[ (4-fluorophenyl)methyl]-1H-benzimidazol-2-yl]carbonyl]-lpiperidinyll-1-butanone (10). A stirred mixture of 8 (5.00 g, 11.1mmol), 9 (6.20 g, 26.0 mmol), KHC03 (2.5 g, 25 mmol), KI (0.19 g, 1.1mmol), toluene (50 mL), and HzO (5 mL) was refluxed 72 h. The reaction was poured into HzO,and the aqueous mixture was extracted with EtOAc. The EtOAc layer was washed with H 2 0 and saturated aqueous NaCl before being dried over an-
10
hydrous MgS04. The drying agent was removed by filtration, and the filtrate was evaporated at reduced pressure leaving an oil. Purification of this oil by flash chromatography (EtOAc) and crystallization from cyclohexane/hexane afforded colorless matted needles: 3.4 g (57%);mp 105-106 "C; IR (KBr) 1690,1670 cm-I; lH NMR (CDC13)6 1.34 (9, 9 H), 1.63-1.80 (m, 2 H), 1.87-2.01 (m, 4 H), 2.16 (dt, J = 11.8, 2.4 Hz, 2 H), 2.43 (t,J = 7.3 Hz, 2 H), 2.92-3.03 (m, 4 H), 3.85 (tt, J = 11.6, 3.8 Hz, 1 H), 5.79 (s, 2 H), 6.90-7.00 (m, 2 H), 7.07-7.15 (m, 2 H), 7.34-7.50 (m, 5 H), 7.89-7.94 (m, 3 H). Anal. Calcd for C&Hs8FN302: C, 75.67; H, 7.10; N, 7.79. Found: C, 75.65; H, 7.16; N, 7.78. Registry No. 2, 124461-07-0; 3, 51-17-2; 4, 124443-67-0; 5, 124461-08-1;6, 124443-68-1; 7, 124443-69-2; 8, 124443-71-6; 9, 43076-61-5; 10,124443-72-7;methyl 4-piperidinecarboxylate hydrochloride, 7462-86-4; 4-fluorobenzyl chloride, 352-11-4; 142bromoethyl)-4-methoxybenzene, 14425-64-0.
Direct Polynitroaliphatic Alcohol Addition to Alkenes. 1. Synthesis of New 2-Fluoro-2,2-dinitroethyl Acetals and Ethers' Robert E. Cochoy,* Raymond R. McGuire,*V2 and Scott A. Shackelfords Frank J. Seiler Research Laboratory (AFSC), USAF Academy, Colorado 80840-6528 Received July 31, 1989
Acetal- and ether-compounds containing the 2-fluoro2,2-dinitroethoxy structure represent an important class of energetic compounds for potential use in formulated propellant and explosive materials, but, their synthesis routes are severely limited. This limitation results from the inherent instability of the gem-2,2-dinitroaliphatic alcohol structure in alkaline or acidic solution4 and from the very weak nucleophilic properties exhibited by this class of alcohol reactants." Therefore, the usual alkaline or acidic conditions for converting alcohols into acetals or ethers cannot be used with gem-2,2-dinitroaliphaticalco(FDNEOH)because hols like 2-fluoro-2,2-dinitroethanol deformylation occurs, producing formaldehyde and either (1) McGuire, R. R.; Cochoy, R. E.; Shackelford, S. A. US Patent 4,424,398, Jan 3, 1984. (2) Present address: Lawrence Livermore National Laboratory, P.O. Box 808 (Code L-282), Livermore, CA 94550. (3) To whom correspondence should be sent. (4) Coburn, M. D. synthesis 1977,570. (5) Cochoy, R. E.; McGuire, R. R. J. Org. Chem. 1972,37,3041-3042. (6) Shipp, K. G.; Hill, M. E., Kamlet, M. J. NOLTR 62-68, April 1962. (7) Adoiph, H. G.; Kamlet, M. J. J. Org.Chem. 1968, 34, 45-50.
This article not subject to U S . Copyright. Published 1990 by the American Chemical Society
1402
J. Org. Chem., Vol. 55, No. 4, 1990
Notes
Table I. 2-Fluoro-2,2-dinitrothanol Catalytic Addition to Unsaturated Hydrocarbons reactant
solvent
HC E COCHzCH3
CHZCI,
prod. no.
products OFONE I H3C -COCH,CH,
catalyst
1
I
Hg(OCOCH3) 2
% yield
prod. distribution
95
-2 FOUND ONLY
OFONE
ONCE I N A MIXTURE OF 1 (73%) A N 0
AN0 H,C=
0
CCI4
OFDNE I COCHZCH,
O O F o N E
-
-2 -3
---
100
1
HgS04
53
-2 (27%) --_
OFONE
I
H,C=
CHOCH2CH3
CHZCIZ
H3C - CHOCH2CH3
H,C=
CHOFDNE
CHZCI,
OFONE I H3C- CHOFONE
CH2CIz
H3C - CCHZCHZCH,
FH3
H,C=
y
CCHZCHZCH,
3
I
OFONE
H,C=
CHCH= CHZ
CCI4
OFONE I H 2 C = CHCH - CH3 OFONE H3CCH'
-7 (77%)
AND
I
AN0
-8 (23%)
8
CHCH2 OFONE I
HZC= CHOCH = C H Z
OFONE
f
- OCHZCF I
CH2CIZ
H2C = CHOCH
-
CH3
9
HgS04
SEE TABLE Ill
70
OFONE I H3C - CHOCH CH3 I OFDNE ~
the 2-fluoro-2,2-dinitromethyl anion (FC(NO,),-) or 2fluoro-2,2-dinitromethane (FC(NO,),H). In spite of this chemical instability, some 2-fluoro-2,2-dinitroethoxy-substituted esters, formals, ethers, amines, and oximes have been synthesized using FDNEOH;"13 synthesis strategies included Michael additions, Mannich condensations, and trifluoroacetic anhydride condensation. Indirect approaches using trifluoromethanesulfonate (triflate) ester intermediates recently expanded polynitroaliphatic ether ~ y n t h e s e s ~and ~ - ' ~sometimes permit a one-pot procedure. This paper reports the first direct, one-step addition of FDNEOH to certain unsaturated hydrocarbons to form acetal and ether compounds new 2-fluoro-2,2-dinitroethoxy in high yield. These nonaqueous mercury salt catalyzed Markovnikov additions with FDNEOH are achieved under the mild, neutral reaction conditions. The scope and limitations of this new polynitroaliphatic reaction as a complementary alternative to other polynitroaliphatic syntheses are discussed. Results and Discussion Direct addition of weakly nucleophilic 2-fluoro-2,2-dinitroethanol (FDNEOH) to unsymmetrically substituted hydrocarbons by mercury salt catalysis proved to be a general one-step method when the carbon-carbon double (8) Grakauskas, V.; Baum, K. J. Org. Chem. 1969, 34, 3927-3930. (9) Adolph, H. G. J. Org. Chem. 1971, 36, 806-809. (10) Grakauskas, V. J. Org. Chem. 1972,38, 2999-3004. (11)Beard, C. D.;Baum, K.; Grakauskas, V. J. Org. Chem. 1973,38, 3673-3677. ... - ... . . (12) Chapman, R. D.;Andreshak, J. L.; Shackelford, S. A. Chem. 1988,53, 3771-3775.
Table 11. Alkynes/Alkenes Resistant to 2-Fluoro-2,2-dinitroethanol Catalytic Addition at the Milder Ambient Pressure Conditions HI HzC=CCECH
H,C=CC=CH
#C=
N+CHzCZ
HOCH2C
HOCC= CCOH
CHl3
y 3
CCH2OH
CH
f l s
0 II
H,C=
CHOCCF,
H2C= CH(CH212CH3
H2C=CHBr
or triple bond is activated by an electron-donating atom or substituent. Mercury(I1) sulfate, mercury(1) sulfate, mercury(I1) acetate, red mercury(I1) oxide, and phenylmercury(I1) chloride all catalyzed this addition reaction, but generally, the two sulfate salts produced the best results. While this reaction system is similar to the mercury-catalyzed solvomercuration-demercuration olefin additions reported for the more common aliphatic alcohols,14 it differs because the poorly nucleophilic FDNEOH forms Markovnikov acetal and ether derivatives in CH2C12or CCll solvent without requiring an alkaline sodium boro-
J. Org.
(13) Chapman, R.D.;Andreshak, J. L.; Herrlinger, S. P.; Shackelford, S. A.; Hildreth, R. A.; Smith, J. P. J. Org. Chem. 1986,51, 3792-3798.
(14) (a) Brown, H. C.; Geoghegan, P. J., Jr., J. Am. Chem. SOC.1967, 89, 5646-5647. (b) Brown, H. C.; Kurek, J. T.; Rei, M. H.; Thompson, K. L. J. Org. Chem. 1985,50, 1171-1174, and references therein.
J. Org. Chem., Vol. 55, No. 4, 1990 1403
Notes hydride reduction of the mercurial intermediate. Reaction conditions usually require an overnight reflux in CHzClQ or CC14for 16 to 19 h, followed by a simple filtration and in vacuo solvent removal. High yields of the desired fluorodinitroethoxy-substitutedproducts are obtained. As discussed later, addition to butadiene required more severe reaction conditions. Table I displays the unsaturated hydrocarbon substrates, reaction conditions, and yields achieved with this simple reaction procedure. Together, Tables I and 11permit a comparison of the subtle structural variations that determine whether this mercury salt catalyzed FDNEOH Markovnikov addition occurs under the milder reaction conditions outlined. Mercury(I1) salts form a Hg2+.rr complex with an alkyne’s triple bond and permit nucleophilic Sn2-like attack.15 In the presence of mercury(I1) acetate catalyst, ethoxyacetylene readily added 2 mol of FDNEOH even at room temperature to form the bis(2-fluoro-2,2-dinitroethyl)ethoxy ethyl orthoester product 1 in 95% yield (Table I). During the first reaction attempt, ethoxyacetylene was placed into the CH2C12solvent as a 62% hexane solution, and two addition products resulted. The higher boiling orthoester diadduct 1 formed as 73% of the product mixture, while the lower boiling vacuum-distilled product represented 27% of the products and proved to be the ethoxy 2-fluoro-2,2-dinitroethylvinyl acetal 2. Subsequent systematic attempts to produce monoadduct product 2 were unsuccessful (eq 1). Such behavior contrasts with NO2 HC
COCHZCH,
I
+ FCCHzOH I
HgOAcz CHZCI,
NO2
-. YO2
I
I H z C = COCHzCH31 *H3C-COCHzCH3
(1)
OCH,CF
I NO2
2
1
a recently reported mercury(I1) chloride catalyzed oxidative methoxymercuration between either a monoalkyl or monoalkoxy acetylene with the highly nucleophilic alcohol, methanol, which adds only once to form methoxy vinyl ethers or methoxyacrolein acetal products, respectively.16 Apparently, the Hg2+ .rr complex of the unsymmetrical acetal intermediate 2 is vinyl 2-fluoro-2,2-dinitroethoxy highly reactive and is not easily intercepted prior to Sn2 attack by a second FDNEOH molecule. Phenylmercury(11)chloride also produced the ethoxyacetylene orthoester diadduct 1 in a lower (43%) yield, after stirring at room temperature in CH2C12for 21 h. Other less reactive acetylenic compounds without an electron-donating alkoxy substituent at the acetylenic carbon fail to add FDNEOH. Apparently, an electron-donating substituent or oxygenatom-containinggroup is needed at the unsaturated carbon position to form a sufficiently stabilized acetylenic or olefinic mercury complex that is reactive enough for attack by the weak FDNEOH nucleophile. (15) March, J. Aduanced Organic Chemistry, 3rd ed.; Wiley & Sons: New York, p 683. (16)Barlenga, J.; Anzar, F.; Bayod, M. Synthesis 1988, 144-146.
Monofunctional alkenes also add FDNEOH under these mild reaction conditions whenever their localized olefinic bond is substituted unsymmetrically with an electrondonating entity. The slight electron-donating properties of the methyl group are sufficient to cause FDNEOH addition with 2-methyl-1-pentene to form product 6 (Table I), but the absence of a 2-substituted methyl in 1-pentene (Table 11) renders this alkene inert to FDNEOH attack. The methyl group activation is offset when the localized double bond’s electron density can be delocalized by conjugation with another unsaturated bond. This is shown inertness to FDNEOH adby 2-methyl-1-buten-3-yne’s dition (Table 11). Vinyl ethers possessing an electrondonating alkoxy substituent readily react and produce acetal derivatives. So reactive is 2,4-dihydropyran that no catalyst is required for a quantitative yield of the 2fluoro-2,2-dinitroethoxyacetal adduct 3 when reacted in refluxing CCl,; however, this is the only example found that needs no mercury salt catalysis. While 2,4-dihydropyran and ethyl vinyl ether readily form their corresponding acetal products, 3 and 4, respectively, even the 2-fluoro-2,2-dinitroethoxyethyl vinyl ether reactant 11 produces its bis(2-fluoro-2,2-dinitroethyl) ethyl acetal, 5, in good yield (Table I). In spite of the highly electronegative P-fluoro-2,2-dinitroethoxy(-OCH,C(NO,),F) substituent at the vinyl position of 11, electron donation from the oxygen atom to the adjacent carbon-carbon double bond must form a stabilized Hg2+-olefin complex reactive enough for FDNEOH attack. Although vinyl ether 11 adds FDNEOH, the vinyl trifluoroacetate ester with a highly electronegative OCOCF3 substituent does not. It would appear this ester’s nonreactivity results from the trifluoroacetate group’s oxygen atom at the vinylic position being in a resonant a position with the carbonyl moiety. Under mild ambient pressure CH2C12or CCl, solvent reflux conditions, any vinylic, acetylenic, aromatic, or carbonyl conjugation apparently delocalizes an alkene’s unsaturated site enough that the mercury ion complex is not sufficiently reactive for attack by FDNEOH. Under these same reaction conditions, simple alkenes without an electron-donating vinyl substituent do not add FDNEOH; this includes strained alkenes like norbornene. Under more stringent higher pressure reaction conditions, alkenes without an electron-donating vinyl substituent or with a conjugated olefinic bond possibly c8n react with FDNEOH; butadiene represents one example. Conjugated butadiene and unconjugated divinyl ether are two difunctional alkenes that behave quite differently in their catalytic FDNEOH addition reactions. In the presence of mercury(I1) sulfate catalyst, the weakly nucleophilic alcohol adds once to excess butadiene in CC14 solvent when heated 16 h under pressure at 55 OC in a shaking Paar pressure bottle. A Markovnikov-directed 1,2or l,4-addition (eq 2) occurs, giving two vinyl ether prod-
-7
8
ucts, the 3-(2-fluoro-2,2-dinitroethoxy)but-l-ene 1,2-adduct
1404
Notes
J. Org. Chem., Vol. 55, No. 4, 1990
7 (77%) and 1,4-adduct, 1-(2-fluoro-2,2-dinitroethoxy)but-2-ene, 8 (23%). No diadduct products were found with butadiene; this contrasts with divinyl ether, which produces both monoadduct 9 and diadduct 10 products (eq 3).
p H,C=
CHOCH= CH, -t FCCHZOH
I
OR
HgSO, Hg2S04 CHzCIZ
Table 111. 2,2,2-Fluorodinitroethanol/DivinylEther Monoand Diadduct Product Percentages DVE/FDNEOH catalvst" % 9 % 10 anal. method A 18 82 isoltd wt 0.75 0.93 A 27 73 'H NMR 0.99 A 36 64 GLPC/MS 1.00 B 37 63 'H NMR 2.00 B 68 32 'HNMR 2.00 A 63 37 'HNMR 2.00 A 60 40 GLPC/MS 2.00 A 78 22 GLPC/MS B 59 47 isoltd wt 2.00 a
9
1 4
The addition of FDNEOH to divinyl ether (DVE) proceeds under milder conditions in refluxing CH2C12solvent over 16 h with either mercury(I1) or mercury(1) sulfate catalysis. The product distribution of 9 versus 10 is easily varied by the DVE/FDNEOH stoichiometry used. A detailed GLPC/MS analysis of the crude oil isolate of 2:l DVE/FDNEOH stoichiometry using mercury(I1) sulfate catalyst reveals the increasing retention time-product distribution: 2-fluoro-2,2-dinitroethylvinyl ether (11) (6%), monoadduct 9 (70%),ethyl vinyl ether adduct 4 (3%), bis(2-fluoro-2,2-dinitroethyl) ethyl acetal 5 (l%), and the diadduct 10 (20%). The diadduct 10 appears as two overlapping peaks in the 30-m capillary HP SE-30 column and apparently represents a diastereomeric pair caused by the two asymmetric carbon atoms on each side of the bridging oxygen atom. The diastereomer of lower GLPC retention time predominates (56 to 4490) regardless of the DVE/FDNEOH reactant stoichiometry. This diasterereomeric mixture apparently causes the 'H NMR spectrum of 10 to give a multiplet resembling a pentet for the single hydrogen (6 5.07) on each asymmetric carbon; theoretically, a quartet is expected. The three minor products 4,5, and 11 all result from minor competing transetherification reaction pathways like that shown in eq 4. Under different NO2
NO2
I
I
HzC= CHOCH = CHZ f FCCHzOH
I
Hgo
> H2C= CH - OCH,Cf
I
CHZCIZ
(4)
NO2
NO2
20 - 30%
11 reaction conditions, the competing transetherification illustrated by eq 4 alternatively can become the exclusive reaction pathway and produce polynitroalkyl vinyl ethers by reaction of DVE with various p-substituted 2,2-dinitro a l c ~ h o l s . ' ~ JDVE ~ contains ethanol as a stabilizer (4%) even when purified, and the minor transetherification reaction between DVE and the ethanol stabilizer produces the intermediate ethyl vinyl ether that adds FDNEOH to form 4. Similarly, FDNEOH produces a trace of 11, and some 11 continues adding FDNEOH to yield 5. Such transetherification reactions are known in mercury(I1)catalyzed reactions between vinyl ethers and the more (17) Shackelford. S. A.: McGuire, R. R.; Cochoy, R. E. US Patent 4,426,540, Jan 17, 1984. (18) Shackelford, S. A.; Cochoy, R. E.; Loving, B. A.; McGuire, R. R. FJSRL-TR-77-0018, Dec 1977.
A, HgSOd; B, HgpSO,.
common nucleophilic a l ~ o h o b .Depending ~~ upon reaction stoichiometry, either the monoadduct product 9 or the diadduct 10 selectively predominates. Essentially, a 1:1 DVE/FDNEOH molar ratio with either mercury(1) or mercury(I1) sulfate catalyst produces twice as much diadduct 10 as monoadduct 9, while a 2:l DVE/FDNEOH ratio reverses this distribution. Table I11 provides a more detailed correlation of 9 and 10 product distribution as a function of reactant ratio, catalyst used, and product analysis method. A 3:l DVE/FDNEOH reactant stoichiometry provides no significant increase in the monoadduct 9. While a change from mercury(I1) to mercury(1) sulfate catalysis has little affect upon product distribution of the DVE and FDNEOH addition (Table 111), altering the mercury salt anion displays a marked impact. Mercury(I1) acetate catalyzes ambient temperature addition of FDNEOH to ethoxyacetylene and provides a high yield of the bis(Zfluoro-2,2-dinitroethyl)ethoxy ethyl orthoester 1, but little reaction occurs with DVE. Refluxing 40 h in CHzC12, less than 15% reaction occurs between DVE and FDNEOH with Hg(OAc)2catalyst. Analysis by GLPC/MS reveals only two minor reaction products, the vinyl ether 11 (3%) and the DVE monoadduct 9 (5%), plus unreacted alcohol (85%) and product 4 (7%) from the ethanol stabilizer. No trace of the diasteromeric diadduct 10 was observed even though some monoadduct 9 forms. Phenylmercury(I1) chloride, which produces the orthoester 1 with ethoxyacetylene, also yields no appreciable amount of 9,10, or 11 with DVE and FDNEOH even when reacted 68 h at room temperature followed by 5-h reflux in CH2Cl2 solvent. Catalysis of the reaction between DVE and FDNEOH by red mercury(I1) oxide in refluxing CHzClz produces only a transetherification reaction in which 11 is the exclusive product;18 a 20-30% isolated yield indicates that the reaction does not proceed to completion (eq 4). This incomplete transetherification initially is driven by the elimination of the ethenol structural fragment from the DVE molecule to form ethenol's acetaldehyde tautomer as a reaction byproduct.'8 Reflux beyond 20 h fails to improve the yield of 11 and product degradation occurs. This important reaction can be driven to completion by modifing the reaction conditions and much higher yields of 11 result.17J8 This reaction modification and the scope this novel transetherification are addressed in a future paper. Conclusions Markovnikov-directed additions of the weakly nucleophilic 2-fluoro-2,2-dinitroethanol (FDNEOH) to a variety of reactive unsaturated hydrocarbons is achieved by mercury salt catalysis under neutral, nonaqueous reaction (19) Watanbe, W. H.; Conlon, L. E. J. Am. Chem. SOC.1957, 79, 282a-2833.
J . Org. Chem., Vol. 55, No. 4, 1990 1405
Notes
conditions. This finding expands the scope of potentially available 2-fluoro-2,2-dinitroethoxy acetal and ether compounds that suffer from severe limitations in possible synthesis strategies. Both the product yield or the specific 2-fluoro-2,2-dinitroallroxycompound produced can depend upon the FDNEOH/alkene reactant stoichiometry or the specific mercury salt catalyst used. Under overnight ambient pressure reflux in CH2C12or CCll solvent, unsaturated hydrocarbons must possess an electron-donating substituent or oxygen atom at its C2 or vinylic position. Otherwise, addition of the poor FDNEOH nucleophile fails to occur; but, more stringent reaction conditions in a sealed Paar pressure bottle may extend this reaction to less reactive alkynes and alkenes as shown by butadiene's reaction with FDNEOH. Demonstration of the subject reaction further opens the possibility of producing analogous new energetic acetal and ethers using other weakly nucleophilic 2,2-dinitroalkyl-substitutedalcohols that also are subject to the facile deformylation reaction under alkaline or acidic conditions.
Experimental Section General. The divinyl ether (DVE) used was purchased from PCR, Inc., Gainee.de, FL,20and initially was used without further purification. Later purchases came from Marshallton Research Laboratory, Winston-Salem,NC. In later reactions, the ethanol stabilizer was enriched by DVE evaporation. The ethanol stabilizer was reduced to 4% ('H NMR analpis) by two distillations (7 in. Vigereaux column), the second over CaH,. Other alkenes and alkynes were distilled only when necessary. The 2-fluoro2,2-dinitroethanol (FDNEOH) was purchased from the Naval Surface Warfare Center/ White Oak Laboratory, Silver Spring, MD, as a 30% by-weight solution in CH2C12solvent.21 Prior to use, the CH2C12was removed by rotary evaporation to yield a slightly yellow oil. (Caution! T h e FDNEOH during solvent removal goes through a sensitivity maximum between 30% solution and neat compound; this operation should be conducted behind appropriate shielding.) The viscous FDNEOH was vacuum distilled in a short path column (bp 36.0-38.8 OC at 0.1 mmHg). Caution! FDNEOH can be explosive under the proper stimulus and also causes severe burns to the skin. Proper shielding and skin protection should be used when handling it or when working u p reactions containing this reagent. The distilled FDNEOH always contained a trace of 2-fluoro-2,2-dinitroethyl methyl formal. The mercury(1)and -(II) sulfate catalysts were obtained from J. T. Baker Chemical Co. ('Baker Analyzed" reagent) as was the red mercury(II)oxide. Mercury(II) acetate came from Fisher Scientific Co. (A.C.S. grade). Neutral aluminum oxide used in the purification and workup was 'Baker Analyzed" purity (pH 6.9-7.4 water slurry). The CH2C12and CCb solvents were MCB spectrometric grade. Nuclear magnetic resonance 'H spectra were taken either on a Varian A-60, T-60, or JOEL FXQ9O instrument (CDC13solvent and TMS internal reference). Infrared spectra were obtained as a neat liquid film (NaCl plates) on a Beckman IR-20 spectrophotometer. Mass spectra were taken on a Dupont 21-491 double focusing mass spectrometer or with a Hewlet Packard 5985 GC/MS system. Elemental analyses were performed by Galbraith Laboratories, Knoxville, TN. Addition of 2-Fluoro-2,2-dinitrothanol to Unsaturated Aliphatic Compounds (General Procedure). The alkene/ alkyne was weighed into CH2Clp or CCll solvent contained in an appropriate-sized round-bottom flask with a Teflon-coated stir bar. FDNEOH was added to the stirred solution; then, the solid Hg(1) or -(II) salt catalyst was added. The reaction flask was fitted with a water-cooled reflux condenser topped with a Drierite filled drying tube. The reaction was sometimes stirred at room temperature, but usually required reflux for 16-19 h. Reaction so-
(20) Divinyl ether is no longer available from this source. Later,
quantities were obtained from Marshallton Research Laboratory, P.O.
Box 11646, Winston-Salem, NC 27106.
(21) This material is available from Fluorochem, Inc., 680 S. Ayon Ave., Azusa, CA 91702.
lution filtration and the solvent removal by rotary evaporation followed. The remaining oil was taken up in 1-3 mL of CCI, and was eluted through a short neutral aluminum oxide column to remove unreacted FDNEOH. This column was prepared by packing 2.5 g of aluminum oxide slurried in CC14 into a 15-mL *course" glass sintered Buchner funnel. CCl, removal afforded the addition product(s). Vacuum distillation provided further product purification. Addition of 2-Fluoro-2,2-dinitroethanolto Ethoxyacetylene (Products l and 2). A flask charged with 1.05 g (15 mmol) of distilled ethoxyacetylene,25 mL of CH2C12,3.08 g (20 mmol) of FDNEOH, and 100 mg of mercury(II) acetate was stirred overnight at room temperature (ca. 16 h). Solvent removal and filtration of the resulting oil through 23 g of alumina (pH = 7.2) with CH2C12,followed by CH,C12 removal, afforded 3.59 g (95%) of clear yellow oil 2. Vacuum distillation (molecular still), 90 OC/O.O5 mm for 4 h, gave 3.26 g of leas pure oil: density 1.42 g/&, NMR (dd, 4.64,4 H) with Jub.m= 17 Hz, (q,3.57,2 H), (~,1.54, 3 H), (t, 1.22,3 H); IR (neat film) cm-' 2990,2950,2900 (sat. CHI, 1600,1310 (NO,). Anal. Calcd for C$I12N407F2:C, 25.4; H, 3.20; N, 14.8; F, 10.1). Found: C, 25.4; H, 3.19; N, 14.7; F, 10.1). In one experiment, 27% vinyl acetal 2 was obtained by distillation (48OC/O.2 mm) of the crude oil mixture with 1: NMR (dd, 4.94, 2 H) with Juie-"F= 16.5 HZ (9, 3.96, 2 H), (9 3.37, 2 H), (t, 1.33, 3 H). Addition of 2-Fluoro-22-dinitroethanolto 3,4-Dihydropyran (Product 3). A flask charged with 1.00 g (11.9 mmol) of 3,4-dihydropyran, 10 mL of CCl,, and 1.00 g (6.5 mmol) of FDNEOH was stirred under reflux for 17 h. CCb solvent removal afforded 1.56 g (100%) of light yellow oil: NMR (m, 4.82,l H), (dd, 4.70, 2 H), (m, 3.72, 2 H), (m, 1.68, 6 H). Addition of 2-Fluoro-2,2-dinitrothanol to Ethyl Vinyl Ether (Product 4). A flask charged with 2.15 g (30 mmol) of ethyl vinyl ether, 25 mL of CH2C12,and 3.08 g (20 mmol) of FDNEOH was cooled with stirring in an ice bath before 200 mg of HgS04 was added. The solution was stirred for 16 h at room temperature. Workup gave 4.65 g of crude oil; vacuum distillation through a 6-in. Vigreaux column at 34-35 OC/O.lO mm yielded 3.29 g (73%) of product 4 as a colorless oil: NMR (q,4.92,1 H), (dd, 4.60, 2 H) with Juie-m= 18 Hz, (m, 3.60, 2 H), (m, 1.26, 6 H); IR (neat film) cm-' 2995, 2940, 2900 (sat. CH), 1600, 1315 (NO,); mass spectrum, characteristic m / e 225 (M - l),211,183, 181, 155, 134, 91, 77, 73, 45 (base), 30, 29. Anal. Calcd for C5Hl1N2O6FC, 31.9; H, 4.87; N, 12.4; F, 8.41. Found: C, 31.9; H, 4.69; N, 12.2; F, 8.35. Addition of 2-Fluoro-2,2-dinitrothanol to 2-Fluoro-2,2dinitroethyl Vinyl Ether (Product 5). A flask charged with vinyl ether,ls 20 mL 1.50 g (8.3 mmol) of 2-fluoro-2,2-dinitroethyl of CH2C12,1.28 g (8.3 mmol) of FDNEOH, and 250 mg of HgSO, was stirred under reflux for 24 h. The isolated crude oil was molecular distilled at 68.0-68.4 OC/O.2 mm to yield 1.71 g (61%) of product 5. Colorless oil with a density = 1.55 g/mL: NMR (q,5.02,1 H), (dd, 4.64,4 H) with Jub-m= 17 Hz, (d, 1.43,3 H); IR (cm-') 3000, 2950, 2900 (sat. CH), 1600, 1310 (NO,); mass spectrum, m / e 333 (M- 1),319,181,147,133,91,75,73,57,45, 44,30 (base), 29. Anal. Calcd for C4H8N4O182:C, 21.6; H, 2.40; N, 16.8; F, 11.4. Found: C, 21.6; H, 2.5; N, 16.6; F, 11.2. Addition of 2-Fluoro-2,2-dinitroethanol to 2-Methyl-1pentene (Product 6). (a) A fiask charged with 2.52 g (30 mmol) of 2-methyl-l-pentene, 25 mL of CH2C12,3.08 g (20 mmol) of FDNEOH, and 200 mg of HgSO, was stirred under reflux for 16 h. The purple solution yielded 2.95 g of crude product. Vacuum distillation at 47-48 OC/O.2 mm (&in. Vigreaux column) gave 3.51 g (74%) of nearly colorless oil 6. Redistillation, 45 OC/O.2 mm, was done with a 12-in. glass bead column: NMR (dd, 4.42,2 H), (s amid a mult, 1.17,13 H); IR (cm-I) 2970, 2940, 2880 (sat. CH), 1600,1315(NO,). Anal. Calcd for C7H&O$: C, 40.4; H, 6.36; N, 11.8; F, 7.98. Found: C, 40.4; H, 6.37; N, 11.7; F, 7.76. (b) A flask charged with 1.00 g (12 mmol) of 2-methyl-1pentene, 20 mL of CHPCl,, 1.85 g (12 mmol) of FDNEOH, and 550 mg of Hg$04 was stirred under reflux for 48 h. Workup gave 1.58 g of purple oil (58%)6 NMR (dd, 4.42,2 H), (s amid a mult, 1.17, 14 H).22 (22) This integration should be 13 H; the product possesses a slight unidentified impurity.
1406
J . Org. Chem. 1990,55, 1406-1408
Addition of 2-FIuoro-2,2-dinitroethanol to 1,3-Butadiene (Products 7 and 8). A Paar pressure bottle charged with 40 mL of CCl, was cooled in an ice bath before 4.2 g (7.8 mmol) of 1,3-butadienewas bubbled into the solvent. Next, 3.08 (2.0 mmol) of 2-fluoro-2,2-dinitroethanol and 100 mg of HgS0, were added to the Paar bottle. The bottle, stoppered with a Teflon-brand wrapped rubber stopper, was shaken at 55 "C for 16 h. The reaction product was then washed through 23 g of alumina (pH = 7.2) with CC,. The CCl, was removed; the product was again filtered through 23 g of alumina with a CCl, wash. CCl, removal provided 4.05 g of light reddish brown oil. Vacuum distillation (6-in. Vigreaux column) at 35 OC/O.lO mm gave 2.19 g (53%) of light yellow oil. The distillate contained mainly the 1,2-adduct 7 with some 1,4-adduct 8. Distillation (51.0-51.5 OC/1.6 mm) (12-in. glass bead column) provided nearly pure (90%) 1,e-adduct 7; the 1P-adduct would not distill even with a diethyl succinate pot chaser. Analytical samples of the two adducts were obtained by preparative GLPC (8 ft. by ' / z in. 20% Dow 710 silicon oil column) at 148 "C. IP-Adduct 7: NMR (m, 5.55,3 H), (dd, 4.48, 2 H) with JuicHF = 18 Hz, (pent, 4.02,l H), (d, 1.25,3 H); lR (cm-') 3090 (=CH), 2990,2930,2890 (sat. CH), 1600,1310 (NO,). Anal. Calcd for C5H&,O6F: C, 34.6; H, 4.36; N, 13.5; F, 9.13. Found: C, 34.85; H, 4.39; N, 13.3; F, 9.16. 1,4-Adduct 8: NMR (m, 5.62, 2 H), (dd, 4.49, 2 H) with Jvie.HF = 18 Hz, (d, 1.76, 3 H); IR (cm-') 3010 (=CH), 2985,2960,2920,2870(sat. CHI, 1600,1315 (NO,). Anal. Calcd for C5HSNz06F:C, 34.6; H, 4.36; N, 13.5; F, 9.13. Found: C, 34.85; H, 4.39; N, 13.6; F, 8.71. Addition of 2-FIuoro-2,2-dinitroethanol to Divinyl Ether (Products 9 and 10). (a) A flask charged with 1.05 g (15 m o l ) of divinyl ether (DVE), 25 mL of CHZClz,3.08 g (20 mmol) of FDNEOH, and 200 mg of HgSO, was stirred under reflux for 16 h. Short-path vacuum distillation (43.2-43.4 OC/O.3 mmHg) of the isolated oil afforded 0.4 g (12%) of monoadduct 9. The pot residue was taken up in CCl, and passed through a short alumina (pH = 7.2) column. CCl, removal gave 2.20 g (58%) of pure diadduct 10, density = 1.42 g/mL. Diadduct 10 NMR (pent, . ~18~ Hz, (d, 1.39, 6 H); IR 5.07, 2 H), (dd, 4.57,4 H) with J u i c = (cm-') 3000, 2940 (sat. CH), 1600, 1310 (NOz);mass spectrum, characteristic m / e 181 (higher GLPC diastereomer), 147,133,119, 105,91,75,73,45(base),44,30,29. Anal. Calcd for c6H12N4011F2: C, 25.4; H, 3.20; N, 14.8; F, 10.1. Found: C, 25.6; H, 3.20; N, 14.7; F, 10.0. (b) DVE (2.10 g, 30.0 mmol), FDNEOH (2.31 g, 15.0 mmol), and 200 mg of HgSO, in 25 mL of CHzClzrefluxed 16 h produced 3.28 g of crude oil product. GC/MS analysis revealed the following crude product distribution: 11 (6%), 9 (70%), 4 (3%), 5 (l%), and 10 (20%). Prolonged or gradual heating during distillation causes apparent polymerization of 9. (c) A flask charged with 2.0 g (28.6 mmol) of DVE, 60 mL of CH,Cl,, 2.2 g (14.3 mmol) of FDNEOH, and 750 mg of HgzS04 was stirred under reflux for 26 h. Workup produced 2.63 g of crude oil containing both monoadduct 9 and diadduct 10. Short-path vacuum distillation afforded 1.29 g (40%) of monoadduct 9. The pot residue was dissolved in CH2ClZand eluted through a short alumina column. CHzC12removal gave 0.91 g (34%) of diadduct 11. Monoadduct 9: NMR (dd, 6.36, 1 H), (4,5.14, 1 H), (dd, = 18 Hz, (m, 4.50, 2 H), d, 1.40, 3 H); IR 4.62, 2 H) with Juie.HF (cm-') 3120,3070 (=CH), 3000,2945 (sat. CH), 1645 (C=C), 1600, 1315 (NO,); mass spectrum, characteristic m / e 181,134, 105,91, 87,71,45 (base), 44,43,30,29. Anal. Calcd for C5HON206F:C, 32.2; H, 4.05; N, 12.5, F, 8.48. Found: C, 32.0; H, 3.98; N, 12.5; F, 8.31.
Acknowledgment. We are deeply indebted to Dr. H. G. Adolph and Dr. M. J. Kamlet (deceased Feb 1988), NSWC/WOL, for their helpful technical discussions and encouragement. Mr. J. L. Pflug (FJSRL) provided extensive 'H NMR and GLPC/mass spectral analyses; Dr. Clay M. Sharts (SDSU) and Dr. John E. Marlin (FJSRL) provided helpful manuscript comments, and Mrs. Linda Pukajilo (FJSRL) aided in manuscript preparation. The Air Force Office of Scientific Research through Dr. D. L. Ball, Director of Chemical and Atmospheric Sciences, generously provided financial support. Registry No. 1, 88934-30-9; 2, 124618-93-5;3, 124618-94-6; 0022-3263/90/1955-1406$02.50/0
4, 124618-95-7; 5, 124618-96-8; 6,88934-26-3; 7, 88934-28-5; 8, 88934-29-6; 9, 88934-27-4; (&)-lo, 124618-97-9; meso-10, 124618980; 11,52483-76-8; FDNEOH, 17003-75-7;DVE, 109-93-3; H&=CHOFDNE, 52483-76-8; EtOC*H, 927-80-0;EtOCH= CH,, 109-92-2;HzC=C(Me)CHzCH2CH3,763-29-1;HzC=CHCH=CHz, 106-99-0; Hg(OAc)z, 1600-27-7; HgSO4, 7783-35-9; HgzS04,7783-36-0;HgO, 21908-53-2;3,4-dihydropyran, 110-87-2.
A Short Enantiodivergent Synthesis of the Geissman-Waiss Lactone' Mikkel Thaning and Lars-G. Wistrand*
Division of Organic Chemistry 3, Chemical Center, University of Lund, P.O.Box 124, S-221 00 Lund, Sweden Received September 5, 1989
The Geissman-Waiss lactone (1) has been shown to be a useful intermediate in the synthesis of pyrrolizidine alkaloids such as retronecine.2 In recent years, (+)-l(with absolute stereochemistry as depicted in 1) has been used
1
3
2
in the chirospecific synthesis of a number of pyrrolizidine alkaloids such as (-)-platynecine (2) and (+)-retronecine (3).3 The enantioselective synthesis of (+)-1 from an optically active starting material has been achieved by several groups., We here report a short synthesis of both (+)- and (-)-16 from a common intermediate, thus making it possible to synthesize both enantiomers of various pyrrolizidine alkaloids. We recently reported on the preparation and reactivity of the N-acyliminium ion 5 (generated in situ from the electrochemically prepared a-methoxylated carbamate 4).8 RO,.
R",. Lewis
4d0,Me )-
Me0
I
4b; R = A c C; R = TBDMS
Q
MeJSiCN
R",.
___)
I
dO,Me 5
&O,Me 6
It was shown that, on reaction with Me,SiCN, the stereoselectivity could be controlled t~ give 6 in cktrans ratios varying from 86:14 to 42:58 by using different 0-protective groups. In addition, amidoalkylation occurred without racemization at C(3).'s8 The synthon 5 incorporates the same absolute stereochemistry at C(3) as (+)-l,and in addition, C(2) is easily functionalized via reaction with a nucleophile. Thus, intermediate 5 is an ideal starting point for the enantioselective synthesis of (+)-1. The introduction of the carbon chain at C(2) in 5 was achieved by using allyltrimethylsilane as the nu~leophiles.~ (Scheme I). Control of the stereoselectivity was again possible by using different 0-protective groups. The cis:trans ratios varied from 77:23 (R = TBDMS) to 21:79 (R = Ac). Only very small effects on the stereoselectivity were observed when the temperature and the Lewis acid were changed (see Table I). The separation and identification'O of the different allyl compounds were carried out
* Author to whom correspondence should be addressed. 0 1990 American Chemical Society
